Skip to content

AGENTS.md

Latest commit

 

History

History
30 lines (22 loc) · 1.52 KB

AGENTS.md

File metadata and controls

30 lines (22 loc) · 1.52 KB

Agent Guidelines — automated-intelligence-hol

What This Repo Is

75-minute hands-on lab showcasing Snowflake's end-to-end data platform — from streaming ingestion to AI-powered analytics. Each subdirectory is an independent exercise.

Architecture

setup.sql                      ← Root DDL: creates all shared infrastructure
demos/                         ← Section demos (Cortex AI, data quality, Gen2, Iceberg, interactive, security, SQL features)
dbt-analytics/                 ← dbt project (staging + mart models)
snowflake-cowork/              ← Cortex Agent + Semantic View
snowpipe-streaming-python/     ← Python Snowpipe Streaming SDK
streamlit-dashboard/           ← 7-page Streamlit in Snowflake app

Critical Rules

  1. Each subdirectory is independent. No cross-subdir imports or shared dependencies.
  2. Database: DASH_AUTOMATED_INTELLIGENCE_DB for all exercises.
  3. Role: AUTOMATED_INTELLIGENCE_ADMIN (primary), WEST_COAST_MANAGER (RBAC demo).
  4. Warehouses: HOL_WH (standard), HOL_GEN2_WH (Gen2), HOL_INTERACTIVE_WH (interactive).
  5. Before going public: Squash all commits into a single clean commit before flipping the repo to public.

When Editing

  • Run setup.sql first — it creates all shared objects (tables, dynamic tables, interactive tables, search services, semantic views)
  • Component scripts in subdirectories extend the base setup (e.g., create_agent.sql, setup_west_coast_manager.sql)
  • Do not modify setup.sql unless changing shared infrastructure